Cryobiopsy Outperforms Forceps in Transbronchial Lung Biopsy in RCT

A 1.1-mm cryoprobe used for transbronchial lung biopsy significantly outperformed standard 2.0-mm forceps on diagnostic yield, specimen size, and histologic quality in the first randomized controlled trial to directly compare the 2 approaches.

Jeffrey Thiboutot, MD, MHS, assistant professor of medicine and interventional pulmonologist at Johns Hopkins University School of Medicine in Baltimore, Maryland, was the study's lead investigator and presented its findings at the 2026 American Thoracic Society (ATS) International Conference held in Orlando earlier this month.

The prospective, multicenter randomized controlled trial (RCT) enrolled 490 participants across 9 sites, randomized 1:1 to transbronchial biopsy using either a 1.1-mm cryoprobe (n = 245) or 2.0-mm forceps (n = 245) for 3 indications: pulmonary parenchymal lesions (PPL; n = 192), lung transplant allograft evaluation (n = 249), and diffuse parenchymal lung disease (n = 49). The primary endpoint was overall diagnostic yield assessed by an independent, blinded central pathology core using ATS/ACCP strict criteria for PPL, ISHLT classifiable definitions of acute cellular rejection for lung transplant, and specific histopathologic patterns for diffuse parenchymal lung disease. The composite safety endpoint was bleeding grade ≥3 CTCAE, pneumothorax requiring chest drainage, or respiratory failure within 30 days, overseen by an independent Data Safety and Monitoring Board.

Overall diagnostic yield was significantly higher in the cryoprobe arm than the forceps arm (89% vs 79%; P = .003). Indication-specific analyses showed statistically significant superiority for PPL (83% vs 70%; P = .04) and lung transplant allograft evaluation (96% vs 89%; P = .03). The diffuse parenchymal lung disease subgroup did not reach statistical significance (72% vs 63%; P = .55), a result Thiboutot attributed to underpowering — with only 49 patients in that arm compared with approximately 200 to 250 in the other 2, the absolute difference in diagnostic yield was comparable across all three indications. Specimen quality metrics uniformly favored the cryoprobe: mean specimen area was 24.2 mm² with the cryoprobe versus 13.1 mm² with forceps (P <.001), the Histologic Accessibility Score was 5.6 versus 5.2 (P <.001), and crush artifact was markedly lower at 2.7% versus 8.9% (P <.001). Crush artifact — mechanical distortion of tissue during forceps retrieval — is a recognized source of uninterpretable pathology specimens and a major contributor to the estimated 30% miss rate of conventional transbronchial biopsy.

Regarding safety, no events meeting the composite endpoint occurred in the cryoprobe arm. Four events were recorded in the forceps arm — all pneumothoraces requiring chest drainage — yielding a complication rate of approximately 1.6%, consistent with published forceps biopsy literature. The cryoprobe arm was not powered to demonstrate a statistically significant safety difference, but the absence of any events in that arm was nonetheless notable.

"Cryobiopsy is safe, and it's nothing you need to be worried about — this is not the same procedure that most people think of when they think of cryobiopsy. The traditional cryobiopsies were with much larger probes where you had to remove the entire scope out of the airway. This is a simple procedure done in the same exact method as standard forceps,” he said.

Thiboutot predicted adoption will proceed first through the interventional pulmonology and lung transplant communities before expanding to general pulmonology training programs, with barriers including fellow training curricula, per-use probe cost relative to forceps, and the upfront device investment — topics his group plans to address through a forthcoming cost-efficacy analysis.

Thiboutot’s disclosures include Erbe USA and Verathon.

Reference

Thiboutot J, et al. A randomized controlled trial of cryobiopsy versus forceps for transbronchial lung biopsy. Presented at: American Thoracic Society International Conference; San Francisco, California; May 2026. [Abstract pending publication]